Skip to content

Vacancy-mediated nitrogen diffusion and aggregation via high-fluence electron beam irradiation in HPHT synthesized diamond crystal

FieldValue
JournalJapanese Journal of Applied Physics
Published2026-09-07
AuthorsChikara Shinei, Y. MASUYAMA, Jun Chen, Hiroshi Abe, Masashi Miyakawa
DOI10.35848/1347-4065/aea3d1

The defect balance after high-fluence irradiation assigned about 4.5 ppm of nitrogen, roughly 70% of the initial 6.8 ppm, to defects other than substitutional nitrogen and NV centers. UV photoluminescence identified H3 centers after irradiation at 3.6 × 10¹⁸ e/cm² and annealing at 1375 ± 25 °C for 2 h. The intended conversion of all substitutional nitrogen into NV structures failed. The authors attribute the aggregate formation to vacancy-mediated nitrogen diffusion rather than direct nitrogen diffusion during annealing.

HPHT-grown nitrogen-doped single-crystal diamond contained 6.8 ppm neutral substitutional nitrogen, [Ns⁰]initial. The material retained natural isotope abundances of 99.6% ¹⁴N and 1.1% ¹³C. The authors cut and mirror-polished the crystal parallel to the {111} plane. They laser-isolated one growth sector measuring 2.0 × 1.0 × 0.4 mm³ to avoid averaging defects across several sectors.

Each specimen received 2.0 MeV electron irradiation at room temperature. Total fluence ranged from 5 × 10¹⁷ to 3.6 × 10¹⁸ e/cm². Annealing then ran at 1375 ± 25 °C for 2 h in vacuum. The authors call this combined treatment the e-beam/anneal process.

X-band EPR at about 300 K measured [Ns⁰] and [NV⁻], each with about ±10% concentration error. Confocal PL mapping with 532 nm excitation estimated [NV⁰] from the NV⁰ to NV⁻ intensity ratio, corrected for decay rates and Huang-Rhys factors. UVPL with 325 nm excitation tested for nitrogen-vacancy complexes. Hahn echo measurements used a 2.5 mT field parallel to [111] and fitted the decay with a stretched exponential.

Defect conversion across the fluence series

Section titled “Defect conversion across the fluence series”

In the low-fluence regime, neutral substitutional nitrogen decreased while both NV charge states increased. The [NV⁻] increase rate reached about 0.13 ppm per 10¹⁷ e/cm². The authors compare this with about 0.1 ppm per 10¹⁷ e/cm² in earlier nitrogen-doped diamonds.

From 5 × 10¹⁷ to 1.0 × 10¹⁸ e/cm², total NV concentration increased by 1.5 ppm. Neutral nitrogen fell by 1.8 ppm over the same interval. This near balance indicates that substitutional nitrogen conversion dominated. The estimated concentration of other nitrogen-related defects was nearly zero.

Above 1.0 × 10¹⁸ e/cm², substitutional nitrogen and NV centers declined as fluence increased. The decline was clear above 1.3 × 10¹⁸ e/cm². At 3.6 × 10¹⁸ e/cm², the nitrogen balance assigned about 4.5 ppm to Nother, a group of unidentified nitrogen-related defects.

H3 formation and vacancy-mediated diffusion

Section titled “H3 formation and vacancy-mediated diffusion”

UVPL spectra separated the two defect regimes. The NV⁰ zero-phonon line at 575 nm and its phonon sidebands appeared at both 7 × 10¹⁷ and 3.6 × 10¹⁸ e/cm². The H3 line at 503.5 nm and its phonon sidebands appeared only at 3.6 × 10¹⁸ e/cm². H3, or NVN⁰, is the nitrogen-vacancy aggregation center. Its appearance established H3 as one component of Nother. This formation temperature was lower than the typical reported annealing temperatures above 1600 °C.

PL peak height did not measure the H3 concentration directly. A 325 nm photon has 3.81 eV energy and promotes photo-ionization from NV⁻ to NV⁰. This optical conversion can make the 575 nm line dominate. The authors also list background absorption and non-radiative transfer to nitrogen aggregates as possible causes of weak H3 emission.

The proposed aggregation route differs from conventional nitrogen diffusion. Using a 6.8 eV activation energy for nitrogen displacement by carbon, the authors estimated a diffusion length of about 0.03 nm during the 2 h anneal. Nitrogen atoms were about 6 nm apart at 6.8 ppm. Direct nitrogen diffusion should therefore be negligible.

Published estimates place monovacancies above about 10 ppm after fluence above 1 × 10¹⁸ e/cm². Their estimated spacing was below 3 nm, closer than the 6 nm nitrogen spacing. Vacancy-mediated nitrogen diffusion can increase the nitrogen diffusion coefficient. The authors therefore assign H3 formation to vacancies bringing nitrogen into aggregation positions. This remains an interpretation supported by defect balances and PL, rather than a direct measurement of nitrogen motion.

Electron supply from Ns⁰ to NV⁰ controls the NV charge ratio. The fitted equilibrium constant K was 0.71 ± 0.08 for this sample series. A previous dataset gave 0.53 ± 0.03. The authors considered these values comparable, despite their expectation that acceptor defects would lower K.

Thus, the measured charge balance did not show a major loss of electron transfer from Ns⁰ to NV⁰ when Nother increased. Estimated Ns⁺ concentration also stayed comparable to [NV⁻] at every fluence. The authors link this to strong binding between NV⁻ and Ns⁺ and argue that Nother is unlikely to donate electrons to NV⁰. This result does not identify the remaining Nother structures.

The Hahn echo coherence time was 30 ± 8 μs below 1.0 × 10¹⁸ e/cm². A model using Ns⁰ and NV defects gave 23 ± 2 μs for the same sample. Their agreement indicates that the initial nitrogen concentration controlled T₂ in this regime. Alsid et al. reported about 420 μs below 1.0 × 10¹⁸ e/cm² at 118 ppb nitrogen. The authors’ earlier CVD study reported 26 μs below 5.0 × 10¹⁷ e/cm² at 5 ppm nitrogen.

From 1.3 to 2.6 × 10¹⁸ e/cm², measured T₂ increased as paramagnetic Ns⁰ and NV centers declined. Above 2.6 × 10¹⁸ e/cm², T₂ saturated near 50 μs. At 3.6 × 10¹⁸ e/cm², measured T₂ was much lower than a model containing only Ns⁰ and NV contributions. The discrepancy indicates that paramagnetic Nother defects add a decoherence source.

The best material figure of merit occurred at 1.0 × 10¹⁸ e/cm². This condition combined 1.7 ppm [NV⁻] with T₂ of 30 μs. Applying these values to the optical system of Wolf et al. gave an estimated Hahn echo sensitivity of about 8.5 pT·Hz⁻¹ᐟ². That system had an 8.5 × 10⁻⁴ mm³ excitation volume and 0.65 collection efficiency. It reported 9 pT·Hz⁻¹ᐟ² with about 0.9 ppm [NV⁻] and 50 μs T₂. The estimate assumed equal ODMR contrast.

An ideal XY16-16 sequence with 256 pulses was projected to extend T₂ to about 1.2 ms for the 1.0 × 10¹⁸ e/cm² sample. The same model gave an AC sensitivity of about 1.3 pT·Hz⁻¹ᐟ². These are extrapolations, not measurements from the irradiated sample.

Engineering implications and unresolved questions

Section titled “Engineering implications and unresolved questions”

For NV magnetometry, maximizing electron fluence is not supported by these data. For this sample, the useful operating point was near 1.0 × 10¹⁸ e/cm², where [NV⁻] × T₂ peaked. Higher fluence reduced NV⁻ concentration and introduced defects that limited T₂. Material selection should therefore use EPR, PL, and coherence measurements together rather than substitutional nitrogen conversion alone.

The H3-rich, P1-poor material may serve a different purpose. The authors cite a commercial Ib diamond with less than 200 ppm nitrogen and less than 10% transmittance. Nitrogen absorption lowered detected photon counts for alpha and beta measurements. They expect H3-rich, P1-poor HPHT diamond to improve scintillator performance, but they did not measure scintillator output in this study.

Two limitations remain. First, H3 was identified as one origin of Nother, but the other structures and their magnetic activity remain unresolved. Second, defect formation can depend on synthesis route. CVD diamond can contain vacancy-related NVH complexes even as grown, and equivalent high-fluence behavior in CVD material was not tested.

The natural 1.1% ¹³C abundance limits Ramsey T₂* to about 1 μs. The authors identify ¹²C-enriched HPHT diamond as necessary for improving DC magnetic sensing. Spin-to-charge conversion could also use the NV⁰ population created at high fluence to improve readout contrast, although that protocol was not tested here.

View Original Abstract

The negatively charged nitrogen vacancy (NV − ) center in diamond is a promising point defect for highly sensitive quantum sensing. The formation of high-density NV − centers is essential for improving sensitivity. We performed room-temperature electron beam irradiation (EBI) and annealing on nitrogen-doped high-pressure high-temperature diamond crystals, aiming to convert all substitutional nitrogen into NV structures by increasing the EBI fluence. While the N s 0 to NV 0/− conversion process dominated at low EBI fluences, in a high-fluence region, NV 0/− center and N s 0/+ concentrations decreased with increasing EBI fluence, indicating the formation of unknown nitrogen-related defects such as the H3 center, which is a nitrogen and vacancy aggregation defect. Although the H3 centers were observed at high EBI fluence, their annealing temperature of 1375 ± 25 °C was lower than the typically reported temperatures over 1600 °C. We attribute this low-temperature formation to vacancy-mediated nitrogen diffusion and aggregation.